Until five years ago, civilian and military uses of nuclear energy were by law kept separate in the United States. In 1998, however, Department of Energy (DOE) secretary Bill Richardson decided to permit commercial nuclear power plants to produce tritium for use in nuclear weapons. This controversial decision is the focus of a sobering book by physicist Kenneth Ber-geron, who worked for twenty-five years at Sandia National Laboratory in Albuquerque, New Mexico. Tritium on Ice is a fascinating and frightening insider’s look at the U.S. nuclear establishment, a must-read for anyone concerned with the uses and abuses of nuclear energy.

Tritium, a radioactive isotope of hydrogen that has a single proton and two neutrons in its nucleus, is used in all current U.S. nuclear weapons. Since tritium decays rapidly (its half-life is twelve years), weapons must be replenished at regular intervals. But in the 1990s public revelations of severe safety lapses and mismanagement at DOE facilities across the nation led to a shutdown of aging tritium production reactors. A search began for alternate means of procuring tritium.

Bergeron finds the U.S. government’s decision to reverse long-standing policy and use civilian facilities for tritium production worrisome for many reasons: it runs counter to the spirit of the Nuclear Nonproliferation Treaty; it sets a bad precedent; it provides other countries with a rationale for using their own civilian reactors for nuclear-weapons production. Furthermore, he points out, commercial power plants are more vulnerable to infiltration by spies or thieves.

Worse still, the plants the energy department selected are run by the Tennessee Valley Authority (TVA), which has an abysmal history of safety violations and of persecuting whistleblowers. The design of the TVA containment buildings is particularly bad. These enormous, superstrong concrete and steel structures are designed to prevent the spewing of radioactive waste from the reactor into the outside environment during a core meltdown. The TVA plants, however, rely on obsolete ice condensers (hence the book’s title). These arrays of ice-filled wire baskets suspended above the reactor are supposed to absorb the heat and steam released during a meltdown, but as Bergeron demonstrates, this Westinghouse design is flawed. Extensive studies have shown that it could, in fact, lead to Chernobyl-like disasters. Bergeron argues that tritium production will increase the likelihood of accidents at these reactors by complicating their operations. He also reasons that the secrecy surrounding tritium production will inhibit the free and open discussion needed to ensure that the power plants operate safely.

What Bergeron finds most outrageous is that no new tritium will even be needed for another ten to twenty years. There is more than enough tritium available from dismantled Cold War weapons to replenish those remaining in the arsenal. Still, the nuclear establishment won’t be dissuaded from rushing ahead with production.

Meticulously researched, Tritium on Ice is a challenging yet rewarding read. (ForeWord magazine named it the best political science book to appear in 2002.) Some of Bergeron’s most interesting insights appear in extensive endnotes and appendices. Here we see how Sandia Labs pushed out David Williams, its chief technical expert on nuclear reactor containment, after he questioned a study of a particularly severe type of core meltdown. Only occasionally is a technical term or concept introduced before it is explained, but inescapable in this strange world of nuclear energy is an alphabet soup of acronyms. Fortunately an appendix decodes them.

As Tritium on Ice went to press, the Nuclear Regulatory Commission (NRC) had yet to decide whether or not to approve the TVA’s request to amend its operating license to produce tritium. Unfortunately that has now changed, as Bergeron told me recently: the NRC agreed to the TVA’s request last September. And, Bergeron says, “DOE plans to load Watts Bar with TPBARs [tritium-producing burnable absorber rods] in the fall of 2003.”

In Tritium, Bergeron quotes physicist Alvin Weinberg, who wrote in 1994, “nuclear energy will always be a demanding technology, one that requires a level of technical sophistication that American utilities generally do not possess.” Bergeron’s book made me think no human organization, including the U.S. government, is sophisticated enough to handle it.